We live in unprecedented times and for retailers this challenging environment throws a new curve ball each week. Shopper behaviour is not easy to predict in normal circumstances and Covid-19 has exacerbated this challenge. This can, for example, be seen in foot traffic to stores – a key indicator of shopper behaviour – which has declined from 4% in February 2020 to – 45% in April 2020 and then -13% to -16% in the final four months of 2020.1 These figures illustrate the radical swing in foot traffic to stores over a short period of time. Now more than ever, retailers need to be able to access data in real time to operate efficiently and stay in business.
With Easter just around the corner, many retailers will have purchased stock that they may not be able to shift easily. Having advanced analytics in place means the retailer can make informed decisions including, running additional promotions, cross selling or bundling products. Brands, such as ASOS, who can be agile and respond quickly to the changing nature of buyer behaviour are those who have flourished in the pandemic.
Typically, the retail industry will have access to traditional sources of data including Epsos and supply chain data. Many retailers will also have an incredible wealth of high velocity data at their fingertips which is being drawn from array of sources including social media analytics, high performing websites, Apps, plus in store wi-fi data, CCTV and footfall which can also be used in real time analytics. The volume and speed of this data is phenomenal.
Overcoming human barriers
One of the key reasons why many in the retail industry fall behind when it comes to utilising real time data and AI, is a human barrier. Put simply, many large traditional retailers do not have the type of platforms needed to manage high velocity, high volume data. Human beings cannot physically process and ingest large volumes of data from multiple sources at speed, and this is where AI can bridge the gap.
The first step to success therefore is investing in a platform. One approach here is to create an on-premise platform and this can be incredibly expensive. An alternative view, which was adopted by a large online retailer four years ago to great success, is to shift to a cloud-based solution which is a scalable platform. Cloud-based platforms provide a modern approach for processing large volumes of data from multiple sources at speed. A modern, cloud-based platform enables a retailer to be fit for change and large-scale transformation that many are currently undergoing due to the impact of Covid-19.
When rendezvous architecture meets data science
Once you have a cloud-based platform that can ingest and process huge volumes of high-speed data, how do you make sense of it? Some retailers will have in-house data scientists that can look at certain scenarios. However, what is needed is an ability to put in place progressive algorithms and data science models, and then train those models so that AI can do the churn for you.
It may be that you have a strong team that can create data science models blending the unique aspects of your retail specialism but being able to productionise these models to gain valuable insight to make fast paced decisions is often missing.
When a cloud-based platform, good governance and a robust process is in place the question becomes how do you manage that amount of data? Here lies another human barrier, namely a lack of skills and training. Several retailers have recognised this barrier and put in place training initiatives, for instance, the M&S Retail Data Academy was launched to train teams to make sense of this data.
The key learning point for the retail industry here is that it is important to not just lead with technology, but to fully understand and articulate the gaps you may have so that you will be able to put the right structures and initiatives in place. A technologically advanced platform on its own will not be enough.
Partner for success
Covid has taught us that being able to make informed decisions at pace when an epidemic hits is priceless. It is important to find a technology partner and draw on their wealth of experience to secure a modern platform with an ability to process and productionise data, so that you can apply your industry nuances, knowledge, and skill to it.
Many retailers were taken by surprise when Covid-19 hit because they were not able to make fully informed management and operational decisions fast enough, as they did not have a modern platform in place with the right training and technology partner. Agile decision making, therefore, wasn’t possible. On the other hand, retailers that invested in a modem cloud-based platform to ingest and process high volume and high velocity data are winning.
1 Source Retail Dive, article What 6 charts say about the pandemic’s impact on retail, reference Nami Sumida/Retail Dive: data from Placer.ai.
By Nick Baladi, Retail Account Lead at Adatis